Documentation of Genetomorph
How to use the Command Line Version:
How to use Genetomorph in your own program:
Installation
A
description how to install a module is available at the main
page of the ASV Toolbox project.
The line
you have to copy into the toolbox.start file looks like this:
de.uni_leipzig.asv.toolbox.genetoMorph.GenetoMorph
Introduction
Genetomorph uses a genetic
algorithm to find morphologic structure, so called morphemes. Morphemes are
composed by phonemes and building words by it part. The Genetic algorithm uses
the following steps: initialising, evaluation, selection, recombination,
mutation and creating of new individuals until a termination condition.
How
to use the Gui Version
Using
the tool:
The first step is loading
data by clicking the Open File button(see figure 1)You find example
input files in ./example/genetoMorph .
Open File button to open the File open dialog
above. File open dialog to laod data for the
tool.
figure 1
Now define
the parameters of the algorithm(see figure 2).
all individuals Information about the individuals. Start button to start algorithm. Select this to eliminate equal
children. parameters: Sim Cycles: Numbers of the cycles for the
simulation. Mutation Rate: Number between 0 and 1,
ratio used for mutation during the evolution process. Numbers of Children: Defines how many
“Children” are created from a parent.
figure 2
Start the
algorithm repeatedly until you satisfied with the result. You can correct the
result manually directly in the table(see figure 3).
Double-click a cell of the table to
edit the entry. Use | to separate the functional unit from the prefix.
figure 3
You can let
show you the statistics as graph by clicking on Show Statistics(see figure 4).
Statistic for the individuals showing
the fitness in dependency of the generation. Blue is the average case and
red is the best case. Show Statistic button to open a statistic
like above.
figure 4
You can
save you results to file by clicking the Save File button(see figure 5).
Save File button to open the file
dialog above. File dialog for saving the individuals.
figure 5
How
to use the Command Line Version:
Command:
For
starting the command line version of this tool use the following command:
java -Xmx500M -classpath .;./lib/ASV_Genetomorph.jar -Djava.ext.dirs=.;./lib
de.uni_leipzig.asv.toolbox.genetoMorph.GenetoMorphCL -i inputfile [-o
outputfile -nc numberOfChildren -mr mutationrate -sc simulationCycles -ee]
Options:
-i inputfile specify the
absolut path to the input file
-o outputfile specify the absolute path to the output file, if not specified
output will print to screen(default: print to screen)
-nc NumberOfChildren specify the number of children for one simulation cycle,
should be an integer(default: 10)
-mr muatationrate specify the mutation rate for the simulation, should be an
integer(default: 1)
-sc simulationCycles specify the number of simulation cycles for this
run(default 10)
-ee eliminate equals individuals, (default: not choosen)
Examples:
- use input file
./examples/genetoMorph/nomen_short.txt with 15 children per simulation
cycle, mutation rate of 3 and do 5 simulation cycles, output to screen
java -Xmx500M -classpath .; ./lib/ASV_Genetomorph.jar -Djava.ext.dirs=.;./lib de.uni_leipzig.asv.toolbox.genetoMorph.GenetoMorphCL -i ./examples/genetoMorph/nomen_short.txt -nc 15 -mr 3 -sc 5
- use input file
./examples/genetoMorph/training.txt with 30 children per simulation cycle,
mutation rate of 5, eliminate equals and do 50 simulation cycles, output
to file ./examples/genetoMorph/trainig_30_5_50.txt
java -Xmx500M -classpath .; ./lib/ASV_Genetomorph.jar -Djava.ext.dirs=.;./lib de.uni_leipzig.asv.toolbox.genetoMorph.GenetoMorphCL -i ./examples/genetoMorph/training.txt -nc 30 -mr 5 -sc 50 -ee -o ./examples/genetoMorph/trainig_30_5_50.txt
How
to use Genetomorph in your own program:
Classes
and Methods
It is easy to use
Genetomorph for your own program. You only need the 3 classes Individual, Data
and Simulator which you find in the package de.uni_leipzig.asv.toolbox.genetoMorph.
|
class.method/class.field |
description |
|
Data.loadWordData(String
filename) |
load all data from the
given file |
|
Data.loadConfigFile() |
load the configuration
file for the data |
|
String[] Data.originalData |
String array containing
the data from the begining |
|
Simulator.initialize() |
initialised the simulator |
|
Simulator.setEliminateEquals(boolean
ee) |
if ee is true equal
individuals where eleminated |
|
Simulator.setChildNum(int
num) |
set how many children
will be created in every simulation cycle |
|
Simulator.setMutationRate(int
mr) |
set the mutation rate for
the simulation |
|
Simulator.startSim() |
starts one simulation
cycle |
|
Individual
Simulator.getMyIndiviuals() |
returns the individuals
of the simulation |
|
int[]
Individual.getGenome() |
returns and int array
containing the position of the end of the morphemically structure |
Example
Here are an example of a
JAVA class(GenetomorphTest.java) using the Genetomorph tool. You can find the
class GenetomorphTest.java in the package de.uni_leipzig.asv.toolbox.tests.
package
de.uni_leipzig.asv.toolbox.tests;
import java.io.File;
import
de.uni_leipzig.asv.toolbox.genetoMorph.Data;
import
de.uni_leipzig.asv.toolbox.genetoMorph.Individual;
import de.uni_leipzig.asv.toolbox.genetoMorph.Simulator;
public class
GenetomorphTest {
/**
* @param args
*/
public static void main(String[]
args) {
//input file
String
inFile = "./examples/genetoMorph/nomen_short.txt";
Data.loadWordData(new File(inFile));
Data.loadConfigFile();
Simulator
sim = new Simulator();
sim.initialize();
//no equal data
sim.setEleminateEqual(true);
//2 children per simulation cycle
sim.setChildNum(2);
//mutation rate of 0
sim.setMutationRate(0);
//do 3 sim cycles
for(int i = 0; i <
3; i++){
sim.startSim();
}
//original Data
String[]
oriData = Data.originalData;
//got result
Individual
ind = (Individual) sim.getMyIndividuals().get(0);
int[] simresult =
ind.getGenom();
// output result
for(int i = 0; i <
oriData.length; i++){
System.out.print(oriData[i].substring(0,
simresult[i])
+ "|" +
oriData[i].substring(simresult[i])+" ");
}
}
}
You
can start this test. Below you see the output of the test.
Words in data set: 102
Processing config file
(./config/genetoMorph/genetoMorph.config)...
Set prefixPenaltyFunc = 3
Set suffixPenaltyFunc = 3
Set k1 = 1.0
Set k2 = 1.0
Set longSuffixPenalty = 1.0
Set containsBase = false
Generate Individuals...
....................................................................................................
done.
Crossover and Mutation... done.
Calc Fitness... .......... done.
Select next Generation... done.
Crossover and Mutation... done.
Calc Fitness... .......... done.
Select next Generation... done.
Crossover and Mutation... done.
Calc Fitness... .......... done.
Select next Generation... done.
Aal|
Aale| A|alen Abbild| Abbild|er Abbildu|ng Abbildunge|n A|bend Ab|ende A|benden
Abhandlung| Abhandlungen| Abrechnung| Abrechnungen| Absch|ied Abschied|e
Abschieds| Absch|nitt Abschnitt|e Abschni|tten Absc|hnitts Dokumentar|film
Dokumentarfilm|e Dokumentarfilme|n Dokumentarfilme|r Dokumentar|films A|nteil
An|teile Ante|ilen Ante|ils Ti|sch Ti|sche Tischen| Tisc|hes Tischche|n
Fleisc|h Fleisch|e Fleisc|hes Weltkri|eg Weltkri|ege Weltkrieg|en Weltkrie|ges
Weltkriegs| Afrika| Afrikane|r Afrikanern| Afrikani|sche Afrika|s Brie|f
Br|iefe Br|iefen Bri|efes Dein| Dein|e Dein|en Dein|er Dein|es Ereigni|s
Ereigni|sse Ereigni|ssen Ereigni|sses Bi|ld Bil|de Bi|ldchen Bild|er B|ildes
Bilde|rn Bi|ldnis G|escheh Geschehe|n Geschehen|e Gescheh|enen Geschehe|ns
Gesc|hehnis Wi|ld Wildnis| Wilde| Held| Held|en H|eldes Kolle|g Kolle|ge
Ko|llegen Ko|llegin Semina|r Semina|re Seminare|n Sem|inars Europame|ister
Europameister|in Europameist|ers Erwerb| Erwe|rber Erwerbs| Erwerbun|g
Erwerbunge|n Streck|e Str|eckung Streck|en Rec|hner Rechne|rn Rech|ners